Files
landing_page/refer_landing_page/scripts/extract/verify_counts.py
T

82 lines
3.4 KiB
Python

#!/usr/bin/env python3
import sys
import os
import re
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
sys.path.insert(0, os.path.join(BASE_DIR, "refer_landing_page", "scripts", "extract"))
def verify():
from generate_all import parse_members, parse_publications, parse_lectures, parse_standards
active_m, alumni_m = parse_members()
intl_p, dom_p, pat_p = parse_publications()
semesters = parse_lectures()
bodies = parse_standards()
total_pubs = len(intl_p) + len(dom_p) + len(pat_p)
std_docs_count = sum(len(p["documents"]) for b in bodies for p in b["projects"])
print("=== DATA VERIFICATION GATE (AC-07 / AC-08 / AC-10 / AC-11) ===")
print(f"Active Members: {len(active_m)} (Expected: 16)")
print(f"Alumni: {len(alumni_m)} (Expected: 60)")
print(f"International Papers: {len(intl_p)} (Expected: 68)")
print(f"Domestic Papers: {len(dom_p)} (Expected: 80)")
print(f"Patents: {len(pat_p)} (Expected: 75)")
print(f"Total Publications: {total_pubs} (Expected: 223)")
print(f"Lectures Semesters: {len(semesters)} (Expected: 45)")
print(f"Standard Bodies: {len(bodies)} (Expected: 6)")
print(f"Standard Documents: {std_docs_count} (Expected: 44)")
errors = []
if len(active_m) != 16:
errors.append(f"Active Members count mismatch: got {len(active_m)}, expected 16")
if len(alumni_m) != 60:
errors.append(f"Alumni count mismatch: got {len(alumni_m)}, expected 60")
if len(intl_p) != 68:
errors.append(f"International Papers count mismatch: got {len(intl_p)}, expected 68")
if len(dom_p) != 80:
errors.append(f"Domestic Papers count mismatch: got {len(dom_p)}, expected 80")
if len(pat_p) != 75:
errors.append(f"Patents count mismatch: got {len(pat_p)}, expected 75")
if total_pubs != 223:
errors.append(f"Total Publications count mismatch: got {total_pubs}, expected 223")
if len(semesters) != 45:
errors.append(f"Semesters count mismatch: got {len(semesters)}, expected 45")
if len(bodies) != 6:
errors.append(f"Standard Bodies count mismatch: got {len(bodies)}, expected 6")
if std_docs_count != 44:
errors.append(f"Standard Documents count mismatch: got {std_docs_count}, expected 44")
# Date ISO 8601 validation (AC-11)
content_dir = os.path.join(BASE_DIR, "content")
for fname in ["members.ts", "publications.ts", "lectures.ts", "standards.ts"]:
fpath = os.path.join(content_dir, fname)
if not os.path.exists(fpath):
errors.append(f"Missing content file: {fname}")
continue
with open(fpath, "r", encoding="utf-8") as f:
text = f.read()
# Check BOM (AC-10)
if text.startswith('\ufeff'):
errors.append(f"BOM found in {fname}")
# Check dates
dates = re.findall(r'"(publishedAt|applicationAt|registrationAt|graduatedAt)": "([^"]+)"', text)
for k, d in dates:
if not re.match(r'^\d{4}(-\d{2})?(-\d{2})?$', d):
errors.append(f"Non-ISO 8601 date in {fname}: {k}='{d}'")
if errors:
print("\n❌ DATA VERIFICATION FAILED:")
for err in errors:
print(f" - {err}")
sys.exit(1)
else:
print("\n✅ ALL DATA VERIFICATION GATES PASSED CLEAN (Exit Code 0)")
sys.exit(0)
if __name__ == "__main__":
verify()